Research On Constant Speed Control Of Variable-speed Hydraulic Pump-controlled Motor System

Based on project entitled “Hydraulic driving electric power generator system of themobile machine”, constant speed control of the motor in the pump controlled motor system,which consists of a variable displacement pump and a fixed displacement motor, wasstudied. In order to cope with the influence from the sharp changes of the pump speed andload torque, the dynamic surface control, linear quadratic (LQ) optimal control and mixedsensitivity H∞control were deeply studied. The simulation results show that these controlmethods make the pump controlled motor system meet the static and dynamic characteristicrequirements. Contents of this thesis can be summarized as follows:(1) A variable displacement pump, a fixed displacement motor and an alternate current(AC) synchronous generator were chosen according to the technical requirements. Thepump controlled motor system was designed. Based on the pump internal specific structurewith the tapered pistons, the control force and control torque nonlinear models of thevariable cylinder were obtained on the force analyses of the swashplate and plunger.According to the statics and dynamic simulation results and linearization of swashplateangle, the mathematical model of the variable pump swashplate displacement was deduced.(2) The frequency response of displacement closed loop system of the variabledisplacement pump with its amplifier was analyzed and its mathematical model wassimplified. Mathematical model of the pump controlled motor system was deduced. Thestability, controllablity, observability and the effect of the pump speed and load torquevariations were discussed. A PID controller with integral separation was studied to improvedynamic performance of motor speed, specially curb an excessive overshoot.(3) In order to improve the system rejection performance of the time-varying pumpspeed, a dynamic surface controller with load torque observer was studied. According tomultiple closed loop principle of dynamic surface control, a dynamic surface controllerwith the parametric uncertainties and reduced order load torque observer was proposed.Adaptive functions of parameters estimator, stability conditions of controller and observerwere achieved through the system stability analyses. The simulation results show that theproposed dynamic surface controller could improve the system rejection performance of thetime-varying pump speed and reduce the settling time of the motor speed.(4) In order to reduce the transient regulation of the motor speed as the load torque sharply changes, the LQ optimal control with the model reference adaptive controller wasinvestigated. A flow controller based on model reference adaptive control and recursiveleast squares was proposed to eliminate the effects of time-varying pump speed fromsystem mathematical model. With the structure of LQ optimal controller, a load torqueobserver and a parameters estimator were designed based on dynamic equilibrium equationof hydraulic motor and load. And the compensation coefficient was acquired. Thesimulation results show that the proposed LQ optimal controller has better rejectionperformances of the time-varying pump speed and load torque disturbance.(5) In order to improve robustness of parametric uncertainties and measurement noisesof the pump controlled motor system, the Bang-Bang controller based mixed sensitivity H∞control was studied. A time optimal controller for the variable displacement pump flow wasachieved through Bang-Bang control to weaken the effect of the time-varying pump speed.A full order load torque observer was designed and a compensation strategy withdifferential element was proposed to improve dynamic performance of the motor speed.The simulation results show that the proposed mixed sensitivity H∞controller hadsatisfactory rejection performances and good robustness on dealing with the parametricuncertainties and measurement noises.Finally, the performance indexes of the control algorithms mentioned above werecompared. Their robustness of the parametric uncertainties and measurement noises wereanalyzed. Advantages and disadvantages of these algorithms were pointed out. In theconclusion, the main results and innovations in this thesis were summarized and somesuggestions for the future research were proposed.